Trading Airflow + EMR for Temporal + Bauplan: The Mediaset tale

Mediaset, a major European broadcaster, transformed its complex and sluggish data infrastructure into a modern, efficient data stack using Temporal and Bauplan, drastically improving the performance and maintainability of their systems. Previously reliant on a cumbersome setup involving multiple AWS services like EMR, Glue, Athena, Redshift, and Airflow, Mediaset faced challenges with slow feedback loops and significant data processing delays, exemplified by a news dashboard that took an hour to refresh. By transitioning to Temporal for workflow orchestration and Bauplan for serverless data management, Mediaset achieved significant improvements, reducing data refresh times from an hour to five minutes and shortening development cycles from three months to six weeks. This reimagined architecture not only simplified infrastructure by reducing reliance on numerous AWS services but also enabled rapid iteration and resilient workflows, allowing Mediaset to focus on advanced applications such as AI-driven content summarization and real-time ad optimization.